ON THE COLOUJRS OF THICK PLATES. 167
Whatever appearance is presented on a screen may be seen without ar screen, by placing the eye in such a position as to receive the rays, and adapting it to distinct vision of an object at the distance of the screen in its former position. It is found universally that when the image of the luminous point is seen distinctly it is accompanied by a portion, more or less extensive, of a system of coloured rings or bands. In this way the rings may be seen when the image is virtual, in which case they cannot, of course, be thrown on a screen.
In the experiment described in the introduction, in which a small flame is placed in such a position as to coincide with its inverted image, and viewed directly, the rings seen are remarkable for their fixity, appearing like a bodily object surrounding the flame, and having a definite parallax, whether judged of by the motion of the head, or by the convergence of the axes of the two eyes. The same is true of the system of rings formed when the flame is moved sideways out of the position above mentioned. The reason of this fixity is, that inasmuch as the retardation is independent of x and y, a given point of an imaginary plane drawn through the flame perpendicular to the axis of the mirror belongs to a ring of the same order, whatever be the point of the mirror against which it is seen projected.
6. Having investigated the conditions of distinctness, let us now proceed to consider the magnitude and character of the rings. supposing the luminous point to be situated at a distance p from the mirror, and the rings to be thrown on a screen at the same distance, or else viewed in air. In this case c = cf = p ; and if the luminous point be in the axis e = 0, which reduces (5) to
It readily follows from this expression that a system of rings is formed similar to the transmitted rings of the system to which Newton's name is especially attached. The rings in the present case, however, especially when viewed in air, are far more brilliant, and in this respect more resemble the reflected system. If el be the radius of the first bright ring, for which R = X, the length of a wave of light,